Matt Brundage

Multidistrict litigation (MDL) now accounts for a large share of the federal civil docket, and court-appointed plaintiffs’ leadership teams wield extraordinary power over investigation, strategy, and settlement in proceedings ranging from opioids to social media. Critics, borrowing from antitrust, charge that these influential and lucrative roles are controlled by a cartel of repeat-player firms; defenders answer that repeat play reflects the expertise complex cases require. Drawing on a novel dataset capturing every leadership appointment across more than two decades of MDLs, supplemented by extensive interviews with practitioners, this Article offers the first systematic empirical analysis of leadership selection and finds no evidence of cartel-like concentration. The more pressing problem, it argues, is that quasi-public authority of this scale is exercised through opaque criteria and weaker institutional checks than almost any comparable body in American law, reorienting the reform debate from market concentration toward capacity, representation, and accountability.

American courts treat judicial records as presumptively public, and the governing standards for sealing are strict: a litigant must show a compelling interest, and the court must make specific, document-by-document findings. But do trial judges actually enforce that standard? Combining docket data from more than two million federal civil cases with machine learning and careful hand-coding, this Article finds that at least 90% of motions to seal are granted, that most sealing orders do not even cite the governing standard, and that when a standard is invoked it is frequently the wrong one. The results carry sweeping implications for the open-courts principle, public health and safety, and the work of the Advisory Committee on Civil Rules: in a system already defined by the vanishing trial, courts and policymakers must now confront a vanishing record.

People—including researchers—routinely form beliefs about politics from nonrepresentative samples: their friends, what they see on television, what circulates on social media. A large body of evidence shows that many fail to correct for this selection problem and instead reason as if what they observe is representative of the broader world. This review develops a formal typology of such “selection neglect” and traces its consequences for political belief formation, identifying distinct mechanisms—including homophily and a “squeaky wheel” effect—through which unrepresentative environments distort perceptions of opinion, group attitudes, and political reality.

Local governments provide many of the public services that shape Americans’ lives. Their fiscal capacities vary enormously, with low-income Americans concentrated in low-capacity governments that struggle to provide these services. School-finance litigation has substantially narrowed the corresponding disparities in education spending. However, I find that spending on other services is still highly unequal, closely tracking the tax bases of the local governments that provide them.

This Article argues that these inequalities generate a distinct democratic harm that I call vote devaluation. This harm arises when citizens’ formally equal votes govern localities with substantial policymaking authority but deeply unequal capacities to deliver what residents ask of them. The harm sits at the overlooked intersection of (a) local-government and fiscal-federalism scholarship, which documents unequal fiscal capacities, and (b) political-science and election-law scholarship, which critiques how and why government policy better reflects the preferences of the Haves over the Have-Nots.

To identify the conditions under which unequal fiscal capacity devalues the vote, I develop a formal model, and I test its implications through original empirical analyses. Combining local-government financing data with nationally-representative survey data, I find that low-income Americans are uniquely dissatisfied with local government, and that their satisfaction rises sharply with local fiscal capacity. These findings challenge leading democratic defenses of status quo local government arrangements on their own terms. Overall, recognizing the harm of vote devaluation opens a new frontier for election law, creating opportunities for litigation, agenda-setting, and coalition-building around policy reforms that would redistribute political power.

Many White Americans underestimate the scale of racial economic inequality and oppose policies designed to reduce it. This paper offers an explanation rooted in a new empirical finding: because of residential income segregation, most White Americans live in communities where Black–White income inequality is markedly lower than the national gap. Residents of these low-inequality environments tend to underestimate national racial inequality and to express less support for remedial policy, and a preregistered experiment shows that supplying accurate information about the racial income and wealth gaps increases recognition of the problem and boosts support for policies addressing it. Together, the findings show how local informational environments can seed misperceptions of racial inequality—and how straightforward factual corrections can move the attitudes that follow from them.

Erie Railroad Co. v. Tompkins is celebrated as a charter of judicial humility: federal courts sitting in diversity must apply state law, and where state law is unsettled they may only predict how the state’s highest court would rule. This Article argues that the prediction mandate is far from neutral. Drawing on an original dataset of federal diversity decisions across insurance, tort, products-liability, employment, and consumer-protection law, it shows that the mandate systematically biases the development of diversity common law in a conservative, innovation-averse, liability-narrowing direction through a self-reinforcing dynamic we call the “conservative ratchet”: told to forecast change and warned against overreaching, federal courts resolve genuine doubt against plaintiffs seeking novel claims—or decline to decide at all, deferring to state courts that, because of removal and the Class Action Fairness Act, rarely get the chance to weigh in. The result is a triple inversion of Erie’s aims—a doctrine meant to stop federal judges from making law leaves them making it, a doctrine meant to end forum-shopping entrenches a durable pro-defendant forum, and a doctrine meant to return the common law to the states hands them a frozen, federally authored, and systematically more conservative version of their own law.